The 12" pipe probably won't provide much detention, unless it has significant volume and has some type of restriction on the outlet.  Otherwise the flow will pass through with minimal effect.

For a discussion of roof modeling please see www.hydrocad.net/roof.htm

Peter Smart
HydroCAD Software
www.hydrocad.net

Thanks. I followed the guidance recommended for modeling a roof.  Made the roof area a subwatershed routed into a pond with outlet.  Even though I do not have specific numbers for the drains coming of the roof, I entered 30 X 30" horizontal orifices as the roof outlet, and defined the roof storage area, just to see the impact.  I was suprised to see that even a large number of large diameter outlets would still predict significant detention of flow by the roof.  Does this seem right?

It all depends on the details (rainfall depth, intensity, roof area, drains, etc.)  30x30 Seems a little big for a roof drain.  What is the maximum depth that occurs in your roof-top "pond"?

Peter Smart
HydroCAD Software
www.hydrocad.net

Peter, the storm event I'm evaluating creates a ponding dpeth of 0.11' on the roof.  I just received better data on the drain system.  It consists of 16, 6" orfices flush to roof.  The peak discharge was reduced from 16 to 3 cfs.  I was suprised to see so much detention.

Your case demonstates why it often makes sense to model a flat roof as a detention pond.

Peter Smart
HydroCAD Software
www.hydrocad.net

Since you are using rooftop detention, I assume you are located in a region where snow buildup and icing will not be a factor to clog the roof drains?

Mike McCann
McCann Engineering

I've modeled roof top storage before and it's important to remember that the roof isn't perfectly flat.  A ponding depth of 2" at the roofs edge may be 6" over the roof drain due to the slope of the roof.  Assuming you modeled the outlet as orifices the additional head on the outlet will affect the flow rate.

The roof will also have overflow drains or scuppers to serve as emergency overflows should the deign storm be exceeded or the primary drains become clogged.

Another note is that the roof drains are usually designed by the mechanical engineer for a design intensity (read Rational Method) and not a rainfall volume (read SCS methodology).

Why are you trying to determine the detention on the roof?  Is it to meet runoff rates for a permit or are there other reasons?

If you do plan on using the rooftop ponding there are a few things to remember.  
1. Rooftop Mechanical Units present on the roof usually don't respond well to being partially submerged.
2. Coordinate with the structural as to the depth of ponding on the roof as it affects their loading calculations.
3. Local building and plumbing codes may dictate the ammount of ponding that is allowed on rooftops or there may be special provisions that need to be meet to allow it.
4. The type of roofing system may be affected by ponding on the roof, e.g. the manufacturer may not warrenty the product if roof ponding is used.

Thanks everyone.  This is helpful information.  The roof I am trying to model is existing, and I am interested in possible detention effects to determine permit compliance.  It sound like the bottom line is I need better data to really determine what is happening.

As others have mentioned, the pitch of the roof and the design of the inlets can have a significant effect on their performance and the resulting detention.  Given the low water depths on a roof, a slightly recessed inlet will see significantly greater head, resulting in more flow and less detention (standing water).

Peter Smart
HydroCAD Software
www.hydrocad.net